January 5, 2025
Source: Originally published by Z. Feel free to share widely.
Credit: Jérémy Barande (Flickr)
AI is already very nearly everywhere. So far, it is no big deal, right? It assembles data, controls this or that machine, writes and/or edits lots of messages and even articles, and who knows, maybe even a book or two. It produces songs, short films, and of course lots of software. It talks, it watches, it will read a book or watch a video and summarize it for you. It’s a time saver, a pattern revealer. But has it turned life upside down and inside out? Not really—unless you think we have ubiquitous mistrust and pervasive lying in part due to AI though I would say Social Media is more implicated in that mess. In any case, it has technological side effects? So what. The main effects are gloriously exciting. Goodbye cancer. Hello immortality.
A poll of Ivy League college students that asked how many books students surrounded by ivy walls read a year revealed that many of them don’t read books anymore at all. Many doubt if they could even manage to read a whole book. It doesn’t matter much because they prefer to get a summary lickety split. So why was I worried in some earlier essays that I offered on this topic? I play Go with an AI that trounces me unless I opt for it to play less well than it can. Same for chess. Great opponent. So? So far, no big deal. No?
Let’s look down the road a bit. Five years off, ten years, whatever. Consider the top 100 songs of 2030. Suppose eighty five were written by AIs. Not just the lyrics, but also the music. Even played by AI. Maybe ninety five. Great listening. Hooray. Headphones get better too. Al engineers at work. Same for new novels, best sellers, and for my reading glasses too, supposing I still read. We have so much to enjoy. And of course it’s the same for software, in spades. Likewise for architectural designs and new product proposals. And oh yes, do you need to go to the doctor? Looks like a person, talks like a person, but it’s an AI that has fantastic bed-side manners. The thing makes brilliant diagnoses. And it is so pretty/handsome—you choose. Even out in the boonies, a therapist in every home. Surgery galore. Longer life. Fabulous.
Meanwhile, your daughter is in high school with all AI teachers. They are very patient with her. Indeed, from day care on is all AI territory. Your son has an AI life partner. A perfect soulmate. No need for angst. No more breakup songs. Everyone has a convenient AI agent. It plans for us, shops for us, cooks for us, cleans for us, read for us, writes for us, reads for us, talks for us, hears our woes, shares our ecstasies (supposing we have any). You want my week’s plan, what I feel? Have your AI me a call. My AI will answer. It will reply as I would, in my voice, with my personality. With no forgetfulness, no errors. I’m in my den. On my couch. Watching AI entertainment. A happy idiot.
The news gets delivered each day by AIs who tell us all is great. All is wonderful. We have somehow avoided AIs being misused by nefarious profit-seeking owners to lie, cheat, steal, and manipulate (despite that of late, in our time, there is steadily less attention to such concerns). We have also somehow avoided AIs going rogue to do their own preferred deeds no matter the human costs (despite that in our time we rush to give them independent ability to make their own decisions and implement their own plans). AIs provide all kinds of calculations, prognostications, and creations. They help us like we used to help our pets. What’s to worry about? What’s to regulate? After all, we can always pull the plug.
Well, one answer goes back a step and says that most of that can’t happen. AI’s won’t ever do the things I mentioned. But just months back the most frequent basis for such a claim was that we were running out of data to train AIs with. There is only one internet, after all. So there is only so much data. And lots of it is garbage. But then came a brainstorm. Why not generate new data with AIs? What could go wrong? Is it a solution or cyber in breeding? Off we go.
But still comes the next answer. even if they keep developing these AIs are just machines. Life partners? Give me a break. Nonsense. Humans won’t fall for that. Teachers, doctors, agents who we converse with, who we get instructions from, who we get life advice from, who we have dinner with? Come on. You might as well marry a pretty rock. And then they are going to write, sing, paint, and film better than flesh and blood writers, singers, painters, and film-makers? Surely not. It is all industry hype to get investments.
Hang on. An AI interviews you for two hours. Yes, you. Tomorrow. It asks you diverse questions. It ruminates on your answers. It winds up with your personality. It grabs your voice too. Your look. Your feel. Think about that. It can act as you would act in 85% of all cases now, today, already. Is that hype? Maybe, but what if it isn’t? In five years, 98%? Except it answers way faster and has infinitely more facts to offer. More insights. Will that mean it is 120% you?
There are dozens of AI products right now and more arrive monthly. Soon, AIs will design new AIs. Even with humans designing, it is largely trial and error. We don’t really know how the hell AIs do what they do so well or often even at all. But with AIs designing? For the most part, nowadays mainly only tech bros extensively use them. AI’s are still a bit too clumsy for you and me. But soon anyone will be able to use them. Easy peasy. What fun. Just ask and you shall receive. No more frustration. No more pathos. No more tedium. No more, no more.
Yes, of course one possible worry is bad actors who will misuse AI to manipulate millions via lies and fear. Read that again. Not exactly a little worry. Another kind of worry is AIs that themselves not only decide how to fulfill requests we make of them, but also decide how to fulfill requests they make of one another regardless of collateral damage for cats, dogs, and people. Also serious, no? But I think there is a third kind of worry. Cheered on by us, AIs become able to do all that humans do, not only faster but also valued more by humans than when humans do those things. Hey robot, I appreciate you. This proceeds until humans prefer AIs to other humans for getting things done, meeting needs, being entertained, and even for intimacy. It proceeds until humans become steadily less social, less productive, less initiating, less discerning, and less creative, and AI-guided life-like robots become more of all of that. Until, that is, we become less human-like and they become more human-like.
At this point, for evidence I might start listing AI products that already exist. I might describe their capabilities, compare them to one year ago, and report what people in the field say is in the works. And I might detail how their use would reduce our time sent in diverse pursuits, progressively atrophying our capacities regarding those pursuits. Or I could ask an AI to do that research and to write it up for me, even in my wordy style. But I bet you won’t be surprised that I would rather suggest that you assemble whatever evidence you think relevant and think it through for yourself to a wonderful or to a God-awful conclusion. What you need is just a Google search away. You don’t want to search? Too much work? Too disturbing? I get that, but consider the implications of that answer coupled to the rest of this brief article for what it says we won’t want to do five years from now.
P.S. I in no way mean to suggest that AIs that do more and more and faster and faster translation, programming, transporting, constructing, designing, care-taking, teaching, doctoring, cooking, planning, and conversing don’t pose serious problems for employment and economy more broadly. I am instead suggesting that even if we deal with those problems and also with nefarious use problems and also with rogue AI mayhem problems, we may still confront a social relations problem that threatens what it is to be a person. Like social media has done a spirited tap dance on what it means to be friends, to pay attention, and to find truth, AI may elephant tromp on what it means to be a person.
P.S.S. I also don’t mean to suggest that we should drop everything else and become Luddites or even just well informed critics of AI mania who demand extensive AI oversight and restrictions. Nope. Global heating and ecological suicide are still around. So too is a surging fascist trend. And nukes. I am instead suggesting that along with developing a movement of movements to address those undeniable dangers, we should also have an emerging movement of movements address AI dangers.
Ugly possibilities. Is there some way out of here? Is there an orientation, a mind set, some values, a vision or perhaps mixture of visions that taken all together can fuel resistance to all our impending horrors right up to our winning a world in which none of that persists? Apocalypse avoided. A better world attained. I think we need a bit more attention to conceiving such a worthy viable shared mindset, values, and visions to then act in accord with in 2025.
ZNetwork is funded solely through the generosity of its readers. Donate
Michael Albert`s radicalization occurred during the 1960s. His political involvements, starting then and continuing to the present, have ranged from local, regional, and national organizing projects and campaigns to co-founding South End Press, Z Magazine, the Z Media Institute, and ZNet, and to working on all these projects, writing for various publications and publishers, giving public talks, etc. His personal interests, outside the political realm, focus on general science reading (with an emphasis on physics, math, and matters of evolution and cognitive science), computers, mystery and thriller/adventure novels, sea kayaking, and the more sedentary but no less challenging game of GO. Albert is the author of 21 books which include: No Bosses: A New Economy for a Better World; Fanfare for the Future; Remembering Tomorrow; Realizing Hope; and Parecon: Life After Capitalism. Michael is currently host of the podcast Revolution Z and is a Friend of ZNetwork.
Credit: Jérémy Barande (Flickr)
AI is already very nearly everywhere. So far, it is no big deal, right? It assembles data, controls this or that machine, writes and/or edits lots of messages and even articles, and who knows, maybe even a book or two. It produces songs, short films, and of course lots of software. It talks, it watches, it will read a book or watch a video and summarize it for you. It’s a time saver, a pattern revealer. But has it turned life upside down and inside out? Not really—unless you think we have ubiquitous mistrust and pervasive lying in part due to AI though I would say Social Media is more implicated in that mess. In any case, it has technological side effects? So what. The main effects are gloriously exciting. Goodbye cancer. Hello immortality.
A poll of Ivy League college students that asked how many books students surrounded by ivy walls read a year revealed that many of them don’t read books anymore at all. Many doubt if they could even manage to read a whole book. It doesn’t matter much because they prefer to get a summary lickety split. So why was I worried in some earlier essays that I offered on this topic? I play Go with an AI that trounces me unless I opt for it to play less well than it can. Same for chess. Great opponent. So? So far, no big deal. No?
Let’s look down the road a bit. Five years off, ten years, whatever. Consider the top 100 songs of 2030. Suppose eighty five were written by AIs. Not just the lyrics, but also the music. Even played by AI. Maybe ninety five. Great listening. Hooray. Headphones get better too. Al engineers at work. Same for new novels, best sellers, and for my reading glasses too, supposing I still read. We have so much to enjoy. And of course it’s the same for software, in spades. Likewise for architectural designs and new product proposals. And oh yes, do you need to go to the doctor? Looks like a person, talks like a person, but it’s an AI that has fantastic bed-side manners. The thing makes brilliant diagnoses. And it is so pretty/handsome—you choose. Even out in the boonies, a therapist in every home. Surgery galore. Longer life. Fabulous.
Meanwhile, your daughter is in high school with all AI teachers. They are very patient with her. Indeed, from day care on is all AI territory. Your son has an AI life partner. A perfect soulmate. No need for angst. No more breakup songs. Everyone has a convenient AI agent. It plans for us, shops for us, cooks for us, cleans for us, read for us, writes for us, reads for us, talks for us, hears our woes, shares our ecstasies (supposing we have any). You want my week’s plan, what I feel? Have your AI me a call. My AI will answer. It will reply as I would, in my voice, with my personality. With no forgetfulness, no errors. I’m in my den. On my couch. Watching AI entertainment. A happy idiot.
The news gets delivered each day by AIs who tell us all is great. All is wonderful. We have somehow avoided AIs being misused by nefarious profit-seeking owners to lie, cheat, steal, and manipulate (despite that of late, in our time, there is steadily less attention to such concerns). We have also somehow avoided AIs going rogue to do their own preferred deeds no matter the human costs (despite that in our time we rush to give them independent ability to make their own decisions and implement their own plans). AIs provide all kinds of calculations, prognostications, and creations. They help us like we used to help our pets. What’s to worry about? What’s to regulate? After all, we can always pull the plug.
Well, one answer goes back a step and says that most of that can’t happen. AI’s won’t ever do the things I mentioned. But just months back the most frequent basis for such a claim was that we were running out of data to train AIs with. There is only one internet, after all. So there is only so much data. And lots of it is garbage. But then came a brainstorm. Why not generate new data with AIs? What could go wrong? Is it a solution or cyber in breeding? Off we go.
But still comes the next answer. even if they keep developing these AIs are just machines. Life partners? Give me a break. Nonsense. Humans won’t fall for that. Teachers, doctors, agents who we converse with, who we get instructions from, who we get life advice from, who we have dinner with? Come on. You might as well marry a pretty rock. And then they are going to write, sing, paint, and film better than flesh and blood writers, singers, painters, and film-makers? Surely not. It is all industry hype to get investments.
Hang on. An AI interviews you for two hours. Yes, you. Tomorrow. It asks you diverse questions. It ruminates on your answers. It winds up with your personality. It grabs your voice too. Your look. Your feel. Think about that. It can act as you would act in 85% of all cases now, today, already. Is that hype? Maybe, but what if it isn’t? In five years, 98%? Except it answers way faster and has infinitely more facts to offer. More insights. Will that mean it is 120% you?
There are dozens of AI products right now and more arrive monthly. Soon, AIs will design new AIs. Even with humans designing, it is largely trial and error. We don’t really know how the hell AIs do what they do so well or often even at all. But with AIs designing? For the most part, nowadays mainly only tech bros extensively use them. AI’s are still a bit too clumsy for you and me. But soon anyone will be able to use them. Easy peasy. What fun. Just ask and you shall receive. No more frustration. No more pathos. No more tedium. No more, no more.
Yes, of course one possible worry is bad actors who will misuse AI to manipulate millions via lies and fear. Read that again. Not exactly a little worry. Another kind of worry is AIs that themselves not only decide how to fulfill requests we make of them, but also decide how to fulfill requests they make of one another regardless of collateral damage for cats, dogs, and people. Also serious, no? But I think there is a third kind of worry. Cheered on by us, AIs become able to do all that humans do, not only faster but also valued more by humans than when humans do those things. Hey robot, I appreciate you. This proceeds until humans prefer AIs to other humans for getting things done, meeting needs, being entertained, and even for intimacy. It proceeds until humans become steadily less social, less productive, less initiating, less discerning, and less creative, and AI-guided life-like robots become more of all of that. Until, that is, we become less human-like and they become more human-like.
At this point, for evidence I might start listing AI products that already exist. I might describe their capabilities, compare them to one year ago, and report what people in the field say is in the works. And I might detail how their use would reduce our time sent in diverse pursuits, progressively atrophying our capacities regarding those pursuits. Or I could ask an AI to do that research and to write it up for me, even in my wordy style. But I bet you won’t be surprised that I would rather suggest that you assemble whatever evidence you think relevant and think it through for yourself to a wonderful or to a God-awful conclusion. What you need is just a Google search away. You don’t want to search? Too much work? Too disturbing? I get that, but consider the implications of that answer coupled to the rest of this brief article for what it says we won’t want to do five years from now.
P.S. I in no way mean to suggest that AIs that do more and more and faster and faster translation, programming, transporting, constructing, designing, care-taking, teaching, doctoring, cooking, planning, and conversing don’t pose serious problems for employment and economy more broadly. I am instead suggesting that even if we deal with those problems and also with nefarious use problems and also with rogue AI mayhem problems, we may still confront a social relations problem that threatens what it is to be a person. Like social media has done a spirited tap dance on what it means to be friends, to pay attention, and to find truth, AI may elephant tromp on what it means to be a person.
P.S.S. I also don’t mean to suggest that we should drop everything else and become Luddites or even just well informed critics of AI mania who demand extensive AI oversight and restrictions. Nope. Global heating and ecological suicide are still around. So too is a surging fascist trend. And nukes. I am instead suggesting that along with developing a movement of movements to address those undeniable dangers, we should also have an emerging movement of movements address AI dangers.
Ugly possibilities. Is there some way out of here? Is there an orientation, a mind set, some values, a vision or perhaps mixture of visions that taken all together can fuel resistance to all our impending horrors right up to our winning a world in which none of that persists? Apocalypse avoided. A better world attained. I think we need a bit more attention to conceiving such a worthy viable shared mindset, values, and visions to then act in accord with in 2025.
ZNetwork is funded solely through the generosity of its readers. Donate
Michael Albert`s radicalization occurred during the 1960s. His political involvements, starting then and continuing to the present, have ranged from local, regional, and national organizing projects and campaigns to co-founding South End Press, Z Magazine, the Z Media Institute, and ZNet, and to working on all these projects, writing for various publications and publishers, giving public talks, etc. His personal interests, outside the political realm, focus on general science reading (with an emphasis on physics, math, and matters of evolution and cognitive science), computers, mystery and thriller/adventure novels, sea kayaking, and the more sedentary but no less challenging game of GO. Albert is the author of 21 books which include: No Bosses: A New Economy for a Better World; Fanfare for the Future; Remembering Tomorrow; Realizing Hope; and Parecon: Life After Capitalism. Michael is currently host of the podcast Revolution Z and is a Friend of ZNetwork.
No comments:
Post a Comment